Consider the video from the front page of numer.ai. There, two statements are made

1) A hedge fund possesses a lot of expensive financial data [so numer.ai anonymizes this data before releasing it to the public by some elaborate encryption method]

2) The founder of the hedge fund associated with numer.ai invested a million dollars of his own money

What I don't understand exactly is the connection between these statements:

  • Although in the video this is not stated explicitly, it seems to be implied that the founder paid this million for financial data. Is this inference correct ?

  • Secondly, am I correct assuming that hedge funds buy their data from other companies, or do they somehow "generate" their own data ? Could you name any companies that one could buy high-quality financial data from ?

  • Thirdly -assuming my assumption from the previous bullets were correct - paying one million dollars for financial data seems ridiculous. Is it plausible that the founder indeed paid all that for financial data ? I've read that the costs of setting up hedge funds vary greatly, from 20k to 200k, but that still short, by a large margin, of one million. Why do you think numer.ai's founder had to invest this amount of money ?


Your inference in #1 is incorrect. The million dollars he has contributed is going to be part of the assets of the fund. This is common practice and is a way for the founder to express confidence that the fund will make money. He wants you to come up with a model that he can then use to trade those assets. Presumably he will give you some money if he uses your model and it works.

Regarding #2, there are lots of ways of getting data. Sometimes you can buy it directly from the exchange. You can also buy from vendors like tickdata.com. There are lots of such vendors. Since he makes a big deal about saying it's expensive, I'm assuming he is talking about data at relatively high frequency (not daily, which would be cheap). Stock data is still not bad. Complete US data would be a few thousand dollars (maybe 20K at the most). For someone sitting at home with no capital, that's a lot of money, but for a hedge fund it's nothing.

As an institutional investor, your broker will give you a data feed that will provide all prices in real time (but not historically). If he's been in operation a while, he could have just saved the prices as they came out of the pipe. I don't think that's the case here, though, based on how young he is and how little money is involved.

In short, he paid for some data and has "encrypted" it in such a way that he can legally share it for free. Supposedly his method preserves the structure so that you could write a trading model based on the encrypted data and it would work on real data. Once you have a good trading model, you sell it to him and he will use it to trade his million dollars and whatever other money he is able to gather.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.